Practical Data Mining Techniques and Applications

Data mining techniques and algorithms are extensively used to build real-world applications. A practical approach can be applied to data mining techniques to build applications. Once deployed, an application enables the developers to work on the users’ goals and mold the algorithms with respect to users’ perspectives. Practical Data Mining Techniques and Applications focuses on various concepts related to data mining and how these techniques can be used to develop and deploy applications. The book provides a systematic composition of fundamental concepts of data mining blended with practical applications. The aim of this book is to provide access to practical data mining applications and techniques to help readers gain an understanding of data mining in practice. Readers also learn how relevant techniques and algorithms are applied to solve problems and to provide solutions to real-world applications in different domains. This book can help academicians to extend their knowledge of the field as well as their understanding of applications based on different techniques to gain greater insight. It can also help researchers with real-world applications by diving deeper into the domain. Computing science students, application developers, and business professionals may also benefit from this examination of applied data science techniques. By highlighting an overall picture of the field, introducing various mining techniques, and focusing on different applications and research directions using these methods, this book can motivate discussions among academics, researchers, professionals, and students to exchange and develop their views regarding the dynamic field that is data mining. Cover Half Title Title Page Copyright Page Table of Contents Preface Acknowledgments Contributors 1 Introduction to Data Mining References 2 Review of Latent Dirichlet Allocation to Understand Motivations to Share Conspiracy Theory: A Case Study of "Plandemic" During COVID-19 2.1 Literature Review 2.1.1 Conspiracy Theories 2.1.2 Measuring Conspiracy Theories Beliefs 2.1.3 Textual Analysis and Text Mining 2.2 Methods 2.2.1 Data Collection 2.2.2 Executing Latent Dirichlet Allocation 2.2.2.1 Dataset Cleaning 2.2.2.2 Dataset Preprocessing 2.2.2.3 Preparing the Data 2.2.2.4 Determining Topics and Corresponding Tweets for the LDA Model 2.2.3 Thematic Analysis 2.3 Results 2.3.1 Results From LDA Tuning Run 2.3.2 Results From Thematic Analysis 2.4 Discussion 2.4.1 Observable Conspiracy Theories Motives 2.4.2 Limitations 2.4.3 Practical Implications and Conclusions References 3 Near Human-Level Style Transfer 3.1 Introduction 3.2 Methodology 3.3 Pre-Processing 3.4 Feature Extraction Using Transfer Learning 3.5 Performance Parameters 3.6 Content Loss 3.7 Style Loss 3.8 Total Variation Loss 3.9 Optimization 3.10 Super-Resolution 3.11 Results and Implementation Code 4 Semantics-Based Distributed Document Clustering 4.1 Introduction 4.2 Background and Related Work 4.3 Proposed Approach: Semantics-Based Distributed Document Clustering 4.3.1 Dataset Pre-Processing 4.3.2 Document Representation: Ontology-Based VSM 4.3.3 Distributed K-Means and Bisecting K-Means Algorithm for Document Clustering 4.4 Datasets and Experimental Description 4.4.1 Pre-Processed Datasets 4.4.2 Experimental Setup 4.5 Results and Discussion 4.5.1 Test Cases for Stability of Algorithms (Count of Clusters and Stability) 4.5.2 Test Cases for Accuracy/Quality of Syntactic and Semantic Analysis 4.5.3 Test Cases for Clustering Time 4.6 Conclusion and Future Scope References 5 Application of Machine Learning in Disease Prediction 5.1 Introduction 5.2 Literature Review 5.3 System Architecture 5.4 Algorithm 5.5 Dataset 5.6 Results and Discussion 5.7 Conclusion References 6 Federated Machine Learning-Based Bank Customer Churn Prediction 6.1 Introduction 6.2 Related Works 6.3 Dataset 6.4 Experimental Setup 6.5 Proposed Approach 6.6 Results 6.7 Challenges 6.7.1 Costly Communication 6.7.2 Detection of Malicious Clients 6.7.3 Privacy Concern 6.7.4 System Heterogeneity 6.8 Conclusion References 7 Challenges and Avenues in the Sophisticated Health-Care System 7.1 Introduction 7.2 Organization of the Chapters 7.3 The Challenges Faced By Health-Care Systems 7.3.1 Patients Predictions 7.3.2 Electronic Health Records (EHRs) 7.3.3 Real-Time Alerting 7.3.4 Patient Engagement 7.3.5 Less Use of Health Data for Informed Strategic Planning 7.3.6 Lack of Predictive Analytics in Healthcare 7.3.7 Fraud and Lack of Security 7.3.8 Less Integration of Enormous Data with Medical Imaging 7.3.9 Risk & Disease Management 7.3.10 Increase in Suicide & Self-Harm 7.4 The Technology/Methodology Behind Data Mining 7.5 Conclusion References 8 Unusual Social Media Behavior Detection Using Distributed Data Stream Mining 8.1 Introduction 8.2 Related Works 8.2.1 User Behavior Analysis 8.2.2 Social Media Bots 8.2.3 Existing Mechanisms/Methods/Algorithms 8.3 Proposed System 8.3.1 Training Models 8.3.2 Unusual Behavior Detection 8.4 Data Format to Be Used 8.5 Conclusion References 9 Market Basket Analysis Using Distributed Algorithm 9.1 Introduction 9.2 Challenges of DARM 9.3 Literature Work 9.4 Proposed Algorithm: Transaction Reduction Using Enhanced Distributed ARM (TR-EDARM) 9.5 Reduction in Communication Cost Using Efficient Communication in TR-EDARM Algorithm 9.6 Datasets 9.7 Results and Discussion 9.7.1 Experiment 1: Improvement in Execution Time (Comparative Analysis of TR-EDARM with Three Benchmark Algorithms) 9.7.2 Experiment No. 2: Improvement in Communication Cost (Comparative Analysis of CDA, FDM, ODAM, and TR-EDARM Based on Communication Cost) 9.8 Conclusion References 10 Identification of Crime-Prone Areas Using Data Mining Techniques 10.1 Introduction 10.2 Related Work 10.3 Architecture and Working 10.3.1 Data Collection 10.3.2 Pre-Processing 10.3.3 Model Training and Evaluation Metrics 10.4 Experimental Results 10.4.1 K-Nearest Neighbor 10.4.1.1 Implementing PCA with the KNN Classifier Model 10.4.2 Support Vector Machine 10.4.3 K-Means Clustering 10.4.4 Random Forest 10.5 Data Analysis & Visualization 10.5.1 Identification of Crime Hotspots 10.6 Conclusion References 11 Smart Baby Cradle for Infant Soothing and Monitoring 11.1 Introduction 11.2 Literature Study 11.3 Proposed Solution 11.3.1 Baby Monitoring System 11.3.1.1 Data Collection Unit 11.3.1.2 Controller Unit— Raspberry Pi B+ 11.3.1.3 Baby Soothing System 11.3.2 Data Transfer Unit 11.3.3 Data Analysis Unit 11.3.3.1 Cry Detection Module 11.3.4 Mobile Application 11.3.4.1 App Initialization and First Use 11.3.4.2 Function of the App 11.4 Results 11.4.1 Baby Cry Detection 11.4.2 Mobile Application 11.5 Conclusion 11.6 Limitations 11.7 Future Scope References 12 Word-Level Devanagari Text Recognition 12.1 Introduction 12.2 Literature Review 12.3 Proposed System 12.3.1 Text Recognizer Application 12.3.2 Processing Server 12.3.2.1 Pre-Processing 12.3.2.2 Feature Extraction and Recognition 12.4 Implementation 12.4.1 Training 12.4.2 Validation 12.4.3 Testing 12.5 Conclusion and Future Scope References 13 Wall Paint Visualizer Using Panoptic Segmentation 13.1 Introduction 13.2 Related Work 13.3 Methodology 13.3.1 Wall Segmentation 13.3.2 Edge Detection 13.3.3 Colour Replacement 13.3.4 Colour Harmonies 13.4 Results and Discussions 13.5 Conclusion References 14 Fashion Intelligence: An Artificial Intelligence-Based Clothing Fashion Stylist 14.1 Introduction 14.2 Related Work 14.3 Proposed Method 14.3.1 Adaptive Content Generating Preserving Network 14.3.2 Segmentation Generation 14.3.3 Clothing Image Deformation 14.3.4 Fashion Intelligence Via ALIAS Normalization 14.4 Experiments 14.4.1 Dataset 14.4.2 Qualitative Analysis 14.4.3 Quantitative Analysis 14.5 Conclusion References Index